netlink: Add functionality to copy addresses from outer namespace

Similarly to what we've just done with routes, support NL_DUP for
addresses (currently not exposed): nl_addr() can optionally copy
mulitple addresses to the target namespace, by fixing up data from
the dump with appropriate flags and interface index, and repeating
it back to the kernel on the socket opened in the target namespace.

Link-local addresses are not copied: the family is set to AF_UNSPEC,
which means the kernel will ignore them. Same for addresses from a
mismatching address (pre-4.19 kernels without support for
NETLINK_GET_STRICT_CHK).

Ignore IFA_LABEL attributes by changing their type to IFA_UNSPEC,
because in general they will report mismatching names, and we don't
really need to use labels as we already know the interface index.

Signed-off-by: Stefano Brivio <sbrivio@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>

4 files changed, 58 insertions, 24 deletions

diff --git a/conf.c b/conf.c
index 404c7bc..cb1e09a 100644
--- a/conf.c
+++ b/conf.c
@@ -645,8 +645,10 @@ static unsigned int conf_ip4(unsigned int ifi,
 	if (IN4_IS_ADDR_UNSPECIFIED(&ip4->gw))
 		nl_route(NL_GET, ifi, 0, AF_INET, &ip4->gw);
 
-	if (IN4_IS_ADDR_UNSPECIFIED(&ip4->addr))
-		nl_addr(0, ifi, AF_INET, &ip4->addr, &ip4->prefix_len, NULL);
+	if (IN4_IS_ADDR_UNSPECIFIED(&ip4->addr)) {
+		nl_addr(NL_GET, ifi, 0, AF_INET,
+			&ip4->addr, &ip4->prefix_len, NULL);
+	}
 
 	if (!ip4->prefix_len) {
 		in_addr_t addr = ntohl(ip4->addr.s_addr);
@@ -696,7 +698,7 @@ static unsigned int conf_ip6(unsigned int ifi,
 	if (IN6_IS_ADDR_UNSPECIFIED(&ip6->gw))
 		nl_route(NL_GET, ifi, 0, AF_INET6, &ip6->gw);
 
-	nl_addr(0, ifi, AF_INET6,
+	nl_addr(NL_GET, ifi, 0, AF_INET6,
 		IN6_IS_ADDR_UNSPECIFIED(&ip6->addr) ? &ip6->addr : NULL,
 		&prefix_len, &ip6->addr_ll);
 
diff --git a/netlink.c b/netlink.c
index d93ecda..bc5b2bf 100644
--- a/netlink.c
+++ b/netlink.c
@@ -334,19 +334,18 @@ next:
 }
 
 /**
- * nl_addr() - Get/set IP addresses
- * @ns:		Use netlink socket in namespace
- * @ifi:	Interface index
+ * nl_addr() - Get/set/copy IP addresses for given interface and address family
+ * @op:		Requested operation
+ * @ifi:	Interface index in outer network namespace
+ * @ifi_ns:	Interface index in target namespace for NL_SET, NL_DUP
  * @af:		Address family
- * @addr:	Global address to fill if zero, to set if not, ignored if NULL
+ * @addr:	Global address to fill on NL_GET, to set on NL_SET
  * @prefix_len:	Mask or prefix length, set or fetched (for IPv4)
- * @addr_l:	Link-scoped address to fill, NULL if not requested
+ * @addr_l:	Link-scoped address to fill on NL_GET
  */
-void nl_addr(int ns, unsigned int ifi, sa_family_t af,
-	     void *addr, int *prefix_len, void *addr_l)
+void nl_addr(enum nl_op op, unsigned int ifi, unsigned int ifi_ns,
+	     sa_family_t af, void *addr, int *prefix_len, void *addr_l)
 {
-	int set = addr && ((af == AF_INET6 && !IN6_IS_ADDR_UNSPECIFIED(addr)) ||
-			   (af == AF_INET && *(uint32_t *)addr));
 	struct req_t {
 		struct nlmsghdr nlh;
 		struct ifaddrmsg ifa;
@@ -365,23 +364,23 @@ void nl_addr(int ns, unsigned int ifi, sa_family_t af,
 			} a6;
 		} set;
 	} req = {
-		.nlh.nlmsg_type    = set ? RTM_NEWADDR : RTM_GETADDR,
+		.nlh.nlmsg_type    = op == NL_SET ? RTM_NEWADDR : RTM_GETADDR,
 		.nlh.nlmsg_flags   = NLM_F_REQUEST,
 		.nlh.nlmsg_len     = NLMSG_LENGTH(sizeof(struct ifaddrmsg)),
 		.nlh.nlmsg_seq     = nl_seq++,
 
 		.ifa.ifa_family    = af,
 		.ifa.ifa_index     = ifi,
-		.ifa.ifa_prefixlen = *prefix_len,
+		.ifa.ifa_prefixlen = op == NL_SET ? *prefix_len : 0,
 	};
+	ssize_t n, nlmsgs_size;
 	struct ifaddrmsg *ifa;
 	struct nlmsghdr *nh;
 	struct rtattr *rta;
 	char buf[NLBUFSIZ];
-	ssize_t n;
 	size_t na;
 
-	if (set) {
+	if (op == NL_SET) {
 		if (af == AF_INET6) {
 			size_t rta_len = RTA_LENGTH(sizeof(req.set.a6.l));
 
@@ -416,21 +415,47 @@ void nl_addr(int ns, unsigned int ifi, sa_family_t af,
 		req.nlh.nlmsg_flags |= NLM_F_DUMP;
 	}
 
-	if ((n = nl_req(ns, buf, &req, req.nlh.nlmsg_len)) < 0 || set)
+	if ((n = nl_req(op == NL_SET, buf, &req, req.nlh.nlmsg_len)) < 0)
+		return;
+
+	if (op == NL_SET)
 		return;
 
 	nh = (struct nlmsghdr *)buf;
+	nlmsgs_size = n;
+
 	for ( ; NLMSG_OK(nh, n); nh = NLMSG_NEXT(nh, n)) {
 		if (nh->nlmsg_type != RTM_NEWADDR)
 			goto next;
 
+		if (op == NL_DUP) {
+			nh->nlmsg_seq = nl_seq++;
+			nh->nlmsg_pid = 0;
+			nh->nlmsg_flags &= ~NLM_F_DUMP_FILTERED;
+			nh->nlmsg_flags |= NLM_F_REQUEST | NLM_F_ACK |
+					   NLM_F_CREATE;
+		}
+
 		ifa = (struct ifaddrmsg *)NLMSG_DATA(nh);
+
+		if (op == NL_DUP && (ifa->ifa_scope == RT_SCOPE_LINK ||
+				     ifa->ifa_index != ifi)) {
+			ifa->ifa_family = AF_UNSPEC;
+			goto next;
+		}
+
 		if (ifa->ifa_index != ifi)
 			goto next;
 
+		if (op == NL_DUP)
+			ifa->ifa_index = ifi_ns;
+
 		for (rta = IFA_RTA(ifa), na = RTM_PAYLOAD(nh); RTA_OK(rta, na);
 		     rta = RTA_NEXT(rta, na)) {
-			if (rta->rta_type != IFA_ADDRESS)
+			if (op == NL_DUP && rta->rta_type == IFA_LABEL)
+				rta->rta_type = IFA_UNSPEC;
+
+			if (op == NL_DUP || rta->rta_type != IFA_ADDRESS)
 				continue;
 
 			if (af == AF_INET && addr && !*(uint32_t *)addr) {
@@ -451,6 +476,13 @@ next:
 		if (nh->nlmsg_type == NLMSG_DONE)
 			break;
 	}
+
+	if (op == NL_DUP) {
+		char resp[NLBUFSIZ];
+
+		nh = (struct nlmsghdr *)buf;
+		nl_req(1, resp, nh, nlmsgs_size);
+	}
 }
 
 /**
diff --git a/netlink.h b/netlink.h
index 217cf1e..cd0e666 100644
--- a/netlink.h
+++ b/netlink.h
@@ -16,8 +16,8 @@ void nl_sock_init(const struct ctx *c, bool ns);
 unsigned int nl_get_ext_if(sa_family_t af);
 void nl_route(enum nl_op op, unsigned int ifi, unsigned int ifi_ns,
 	      sa_family_t af, void *gw);
-void nl_addr(int ns, unsigned int ifi, sa_family_t af,
-	     void *addr, int *prefix_len, void *addr_l);
+void nl_addr(enum nl_op op, unsigned int ifi, unsigned int ifi_ns,
+	     sa_family_t af, void *addr, int *prefix_len, void *addr_l);
 void nl_link(int ns, unsigned int ifi, void *mac, int up, int mtu);
 
 #endif /* NETLINK_H */
diff --git a/pasta.c b/pasta.c
index c2fb1f4..798aeb6 100644
--- a/pasta.c
+++ b/pasta.c
@@ -278,16 +278,16 @@ void pasta_ns_conf(struct ctx *c)
 		nl_link(1, c->pasta_ifi, c->mac_guest, 1, c->mtu);
 
 		if (c->ifi4) {
-			nl_addr(1, c->pasta_ifi, AF_INET, &c->ip4.addr,
-				&c->ip4.prefix_len, NULL);
+			nl_addr(NL_SET, c->ifi4, c->pasta_ifi, AF_INET,
+				&c->ip4.addr, &c->ip4.prefix_len, NULL);
 			nl_route(op_routes, c->ifi4, c->pasta_ifi, AF_INET,
 				 &c->ip4.gw);
 		}
 
 		if (c->ifi6) {
 			int prefix_len = 64;
-			nl_addr(1, c->pasta_ifi, AF_INET6, &c->ip6.addr,
-				&prefix_len, NULL);
+			nl_addr(NL_SET, c->ifi6, c->pasta_ifi, AF_INET6,
+				&c->ip6.addr, &prefix_len, NULL);
 			nl_route(op_routes, c->ifi6, c->pasta_ifi, AF_INET6,
 				 &c->ip6.gw);
 		}